At 16:23 02/08/2005, Hallam-Baker, Phillip wrote:
> Behalf Of JFC (Jefsey) Morfin
> This is why I suggest the real danger for the IETF is the
> collusion of
> large organisations through external consortia to get a
> market dominance
> through de facto excluding IETF standardisation and IANA
> registry control.
> And this is why I suggest the best way to address it is
> simply to ask for
> the truth, the whole truth.
Hallam,
I still must answer you very cute remark on what one could name "delta
sec". I am giving a lot of thinking. I find it very interesting. So I will
be careful about this one :-)
The problem with this approach is that it becomes self-defeating. The
work of the IETF gets clogged up by individuals whose sole objective is
to block what they see as the encroachments of evil corporations at all
costs.
This is unfortunately what I must do right now. But unfortunately this is
not what I see, it is what they demonstrate.
Even if they can't see the evil globalization scheme immediately
they will block progress anyway just in case. The result is that
corporations that want to get work done either go to other forums or
craft proposals that are so narrowly drafted that they amount to a
rubber stamp.
Except if you can grab a BCP. I am not sure you are actually right. You
certainly know a few cases. I known one before: I actually partly oppose
your company. I gave up as it was my first IETF opposition. Today I see
that it would have been tremendously beneficiary to your company if I had
hold my position. The problem with IETF is there is no architectural common
vision. So you do not know if your rubber stamp is at the proper place.
This is why would prefer to have a good evaluation of all the interests
supporting a proposition. Having to road map, I could at least understand
who supports. If there is a good distribution of support, this is good. If
there is only a commercial, or a political, etc. support: warning. This is
simply some more sophisticated rough consensus evaluation process. Avoiding
consensus by exhaustion organised by affinity groups.
Certainly there are bizare corporations attempting to achieve some sort
of stranglehold. Anyone remember digital convergence and the CueCat?
That type of behavior tends to come from market entrants rather than
established companies. Once you have a stake in the open Internet the
probability of success in a closed 'walled garden' scheme isn't high
enough to be interesting.
Unless you are dominant and want to protect that dominance.
Furthermore the people working for those corporations tend to consider
themselves advocates for and responsible to their customers and their
customer's customers at least as much if not more than their
shareholders.
dominance makes this the same. You have so many customers that their
stability seems to be part of the internet. But dominance in an area can be
defeated by dominance or greassroots effort in an area which looked
orthogonal. The problem is that it may create disruption. Look at Internet
balkanisation.
Sit at the back of the plenary sessions. Watch the number of people
opening up their laptop and starting a telnet session. Less than 5% of
the billion plus Internet users interact with their machine in that way.
The IETF membership is totally unrepresentative of the billion plus
Internet users. Worse still the prevaling attitude is of the 'anyone can
become like us only not quite so skilled' type. Most people don't want
to have to become computer experts.
The IETF does not have a veto over the development of the Internet.
There are plenty of standards organizations to choose from. Nor for that
matter does IANA. All IANA is is a voluntary arrangement that exists
because people choose to recognize it. There is in practice nothing to
stop individuals simply declaring that they will use a particular code
point.
IETF and IANA have a defacto monopoly on the architecture. This
architecture must evoluate for years. This only lead to the question: will
they make it or who will? Two responses today: ITU or grassroots. If
someone believes the ITU is able to do it .... so it is grassroots. But
grassroots is balkanisation, starting by the dominant securing their
dominant territory. And grassroots undermining it.
This has good and bad effect. At this time I have not yet determined the
best way out of IETF.
As a thought experiment consider what happens if someone decides they
want the DNS RR 88 and just goes and uses it. If they succeed and their
standard is used nobody else is going to accept issue of RR #88. And
that is all anyone needs from IANA.
This total lack of control is actually not such a bad thing. It means
that if the International 'Internet Governance' cabal that wants to
capture the IANA were to succeed the success it would not matter very
much.
The IANA time is over. The problem is its consistent replacement.
> This is the only way to obtain open, scalable and
> uniform standards.
Are these the right goals?
Surely meeting the needs of the users should come somewhere in the list.
This is what I think to be the needs of the users. Certainly interested if
you have other root needs?
Uniformity in standards can be a good thing. But there are also
disadvantages to insisting on 'consistency' with what are at this point
quarter century old designs.
This is why I use uniformity. I attach it to "common understanding" of what
the other does. This does not necessarily means consistency at least the
way we use the the word. But I certainly agree we miss terms we agree upon.
I tend to say that the user need coherence, stability, simplicity
internally, what means open, scalable, uniform outside (with no real
overlap between the notions)?
Ten years ago I would have thought that the idea of 'disposable'
standards whose sole purpose was to effect a transition to some other
standard was mad. Today I really don't see any problem with the idea
that you write a spec whose sole purpose is to enable a transition.
Full agreement. But a model would deeply help to know where the various
disposable solutions lead to. Also remember that disposable may stay a long
time. So there should be no synchronisation need in transtion. For example
the whole IPv6 issue is that they did not understand that their current
deployement (2001) is disposable.
It is pretty hard for any standard to get anywhere unless it is 'open'.
It is not exactly in my employer's interest to allow a competitor to
gain such a position. Nor is it in my competitor's interest to allow me
to achieve such a position.
Your employer did not make very much out of the DNS right now when compared
with what it could have done with aliases and MLDN? The problem is not only
open standards but also to be open minded. Again this comes to a model. My
model works for 20 years now. Still wait for someone discovering it, since
they are too smart to accept it :-)
jfc
_______________________________________________
Ietf@xxxxxxxx
https://www1.ietf.org/mailman/listinfo/ietf